🛠 Tech Stack

💼 About This Role

You'll be developing Vision-Language-Action (VLA) models to enable robots to navigate and interact in heavy industry environments. You'll work on bridging multi-modal perception with physical execution using state-of-the-art VLM/VLA frameworks. This role offers a unique opportunity to combine cutting-edge research with real-world deployment on heavy mobile equipment.

🎯 What You'll Do

• Analyze and preprocess multi-modal datasets (video, radar, lidar)
• Build self-supervised dataset generation pipelines for VLA models
• Fine-tune and optimize VLM/VLA models using GenAI frameworks
• Execute training runs and evaluate model performance on industrial data

📋 Requirements

• Currently enrolled in a Master's program in Robotics, CS, EE, or related field
• Strong hands-on experience with Python and PyTorch
• Deep interest in Vision-Language Models for robotics
• Initial project experience in GenAI, Robotics, or Computer Vision

✨ Nice to Have

• Familiarity with HuggingFace, Gemini, or Unsloth
• Experience with prompt optimization or model fine-tuning
• Previous machine learning research (e.g., publications, thesis)

🎁 Benefits & Perks

• 📈 Attractive compensation package and stock options
• 🍻 Beverages on-site and regular social events
• 🔬 Engage with top-tier researchers and engineers
• 🚀 Influence the future of robotic technologies
• 🌍 Relocation assistance to Berlin

🚩 Heads Up

• Internship role requires advanced GenAI and robotics experience typically expected of a full-time researcher